A two-dimensional model for the flapping of an elastic flag under axial flow is described. The vortical wake is accounted for by the shedding of discrete point vortices with unsteady intensity, enforcing the regularity condition at the flag's trailing edge. The stability of the flat state of rest as well as the characteristics of the flapping modes in the periodic regime are compared successfully to existing linear stability and experimental results. An analysis of the flapping regime shows the co-existence of direct kinematic waves, travelling along the flag in the same direction as the imposed flow, and reverse dynamic waves, travelling along the flag upstream from the trailing edge.

Pressure fluctuations measured at the wall of a turbulent boundary layer are analysed using a bi-variate continuous wavelet transform. Cross-wavelet analyses of pressure signals obtained from microphone pairs are performed and a novel post-processing technique aimed at selecting events with strong local-in-time coherence is applied. Probability density functions and conditionally averaged equivalents of Fourier spectral quantities, usually introduced for modelling purposes, are computed. The analysis is conducted for signals obtained at low Mach numbers from two different non-equilibrium turbulent boundary layer experiments. It is found that that the selected events, though statistically independent, exhibit bi-modal statistics while the conditional coherence function coincides with its non-conditional Fourier equivalent. The physical nature of the selected events has been further explored by the computation of ensemble-averaged pressure time signatures and the results have been physically interpreted with the aid of numerical and experimental results from the literature. In both experiments, it has been found that the major physical mechanisms responsible for the observed conditional statistics are represented by sweep-type events which can be ascribed to the effect of streamwise vortices in the near-wall region. More precisely, the wavelet analysis highlights the convection of the selected structures in both cases. Conversely, compressibilty effects could be related to these events only in one case.

In this paper, a method is introduced that allows calculation of an approximate proper orthogonal decomposition (POD) without the need to perform a simulation of the full dynamical system. Our approach is based on an application of the density matrix renormalization group (DMRG) to nonlinear dynamical systems, but has no explicit restriction on the spatial dimension of the model system. The method is not restricted to fluid dynamics. The applicability is exemplified on the incompressible Navier–Stokes equation in two spatial dimensions. Merging of two equal-signed vortices with periodic boundary conditions is considered for low Reynolds numbers Re≤800 using a spectral method. We compare the accuracy of a reduced model, obtained by our method, with that of a reduced model obtained by standard POD. To this end, error functionals for the reductions are evaluated. It is observed that the proposed method is able to find a reduced system that yields comparable or even superior accuracy with respect to standard POD method results.

We study the relaxation of initially segregated scalar mixtures in randomly stirred media, aiming to describe the overall concentration distribution of the mixture, its shape and rate of deformation as it evolves towards uniformity. A stirred scalar mixture can be viewed as a collection of stretched sheets, possibly interacting with each other. We consider a situation in which the interaction between the sheets is enforced by confinement and is the key factor ruling its evolution. It consists of following a mixture relaxing towards uniformity around a fixed average concentration while flowing along a constant cross-section channel. The interaction between the sheets is found to be of a random addition nature in concentration space, leading to concentration distributions that are stable by self-convolution. The resulting scalar field is naturally coarsened at a scale much larger than the dissipation scale. Consequences on the mixture entropy and scalar rate of dissipation are also examined.

We examine steady longitudinal freezing of a two-dimensional single-component free liquid film. In the liquid, there are thermocapillary and volume-change flows as a result of temperature gradients along the film and density change upon solidification. We examine these flows, heat transfer, and interfacial shapes using an asymptotic analysis which is valid for thin films with small aspect ratios. These solutions depend sensitively on contact conditions at the tri-junctions. In particular, when the sum of the angles formed in the solid and liquid phases falls below a critical value, the existence of steady solutions is lost and the liquid film cannot be continuous, suggesting breakage of the film owing to freezing. The solutions are relevant to the freezing of foams of metals or ceramics, materials unaffected by surface active agents.

Moderately favourable pressure gradient turbulent boundary layers are investigated within a theoretical framework based on the unintegrated two-dimensional mean momentum equation. The present theory stems from an observed exchange of balance between terms in the mean momentum equation across different regions of the boundary layer. This exchange of balance leads to the identification of distinct physical layers, unambiguously defined by the predominant mean dynamics active in each layer. Scaling domains congruent with the physical layers are obtained from a multi-scale analysis of the mean momentum equation. Scaling behaviours predicted by the present theory are evaluated using direct measurements of all of the terms in the mean momentum balance for the case of a sink-flow pressure gradient generated in a wind tunnel with a long development length. Measurements also captured the evolution of the turbulent boundary layers from a non-equilibrium state near the wind tunnel entrance towards an equilibrium state further downstream. Salient features of the present multi-scale theory were reproduced in all the experimental data. Under equilibrium conditions, a universal function was found to describe the decay of the Reynolds stress profile in the outer region of the boundary layer. Non-equilibrium effects appeared to be manifest primarily in the outer region, whereas differences in the inner region were attributed solely to Reynolds number effects.

The capillary instability of a liquid thread containing a regular array of spherical particles along the centreline is considered with reference to microencapsulation. The thread interface may be clean or occupied by an insoluble surfactant. The main goal of the analysis is to illustrate the effect of the particle spacing on the growth rate of axisymmetric perturbations and identify the structure of the most unstable modes. A normal-mode linear stability analysis based on Fourier expansions for Stokes flow reveals that, at small particle separations, the interfacial profiles are nearly pure sinusoidal waves whose growth rate is nearly equal to that of a pure thread devoid of particles. Higher harmonics suddenly enter the normal modes for moderate and large particle separations, elevating the growth rates and yielding a stability diagram that consists of a sequence of superposed pure-thread lobes. A complementary numerical stability analysis based on the boundary integral formulation for Stokes flow reveals the strong stabilizing effect of particles whose radius is comparable to the thread radius. Numerical simulations of the finite-amplitude motion based on the boundary integral method demonstrate that thread breakup leads to particles coated with annular layers of different thicknesses.

We consider levitation of an axisymmetric drop of molten glass above a spherical porous mould through which air is injected at a constant velocity. Owing to the viscosity contrast, the float height for a given shape is established on a much shorter time scale than the subsequent deformation of the drop under gravity, surface tension and the underlying lubrication pressure. Equilibrium shapes, in which an internal hydrostatic pressure is coupled to the external lubrication pressure through the total curvature and the Young–Laplace equation, are determined using a numerical continuation scheme. The set of solution branches is surprisingly complicated and shows a rich bifurcation structure in the parameter space (Bo=ρgV2/3/γ, Ca=μav/γ), where Bo is bond number and Ca is capillary number, ρ and V are the drop density and volume, γ the surface tension, μa the air viscosity and v the injection velocity. The linear stability of equilibria is determined using a boundary-integral representation for drop deformation that factors out the rapid vertical adjustment of the float height. The results give good agreement with time-dependent simulations. For sufficiently large Ca there are intervals of Bo for which there are no stable solutions and, as Ca increases, these intervals grow and merge. The region of stability decreases as the mould radius aM increases with an approximate scaling Ca~aM−5, which imposes practical limitations on the use of this geometry for the manufacture of lenses.

The Kelvin wave is the lowest eigenmode of Laplace's tidal equation and is widely observed in both the ocean and the atmosphere. In this work, we neglect mean currents and instead include the full effects of the Earth's sphericity and the wave dispersion it induces. Through a mix of perturbation theory and numerical computations using a Fourier/Newton iteration/continuation method, we show that for sufficiently small amplitude, there are Kelvin travelling waves (cnoidal waves). As the amplitude increases, the branch of travelling waves terminates in a so-called corner wave with a discontinuous first derivative. All waves larger than the corner wave evolve to fronts and break. The singularity is a point singularity in which only the longitudinal derivative is discontinuous. As we solve the nonlinear shallow water equations on the sphere, with increasing ε (‘Lamb's parameter’), dispersion weakens, the amplitude of the corner wave decreases rapidly, and the longitudinal profile of the corner wave narrows dramatically.

A paradox concerning the flight of insects in two-dimensional space is identified: insects maintaining their bodies in a particular position (hovering) cannot, on average, generate hydrodynamic force if the induced flow is temporally periodic and converges to rest at infinity. This paradox is derived by using the far-field representation of periodic flow and the generalized Blasius formula, an exact formula for a force that acts on a moving body, based on the incompressible Navier–Stokes equations. Using this formula, the time-averaged force can be calculated solely in terms of the time-averaged far-field flow. A straightforward calculation represents the averaged force acting on an insect under a uniform flow, −〈V〉, determined by the balance between the hydrodynamic force and an external force such as gravity. The averaged force converges to zero in the limit 〈V〉 → 0, which implies that insects in two-dimensional space cannot hover under any finite external force if the direction of the uniform flow has a component parallel to the external force. This paradox provides insight into the effect of the singular behaviour of the flow around hovering insects: the far-field wake covers the whole space. On the basis of these assumptions, the relationship between this paradox and real insects that actually achieve hovering is discussed.

An analysis of the sound radiated by three turbulent, high-speed jets is conducted using Lighthill's acoustic analogy (Proc. R. Soc. Lond. A, vol. 211, 1952, p. 564). Computed by large eddy simulation the three jets operate at different conditions: a Mach 0.9 cold jet, a Mach 2.0 cold jet and a Mach 1.0 heated jet. The last two jets have the same jet velocity and differ only by temperature. None of the jets exhibit Mach wave characteristics. For these jets the comparison between the Lighthill-predicted sound and the directly computed sound is favourable for all jets and for the two angles (30° and 90°, measured from the downstream jet axis) considered. The momentum (ρuiuj) and the so-called entropy [p − p∞ − a∞2(ρ − ρ∞)] contributions are examined in the acoustic far field. It is found that significant phase cancellation exists between the momentum and entropy components. It is observed that for high-speed jets one cannot consider ρuiuj and (p′ − a∞2ρ′)δij as independent sources. In particular the ρ′ūxūx component of ρuiuj is strongly coupled with the entropy term as a consequence of compressibility and the high jet velocity and not because of a linear sound-generation mechanism. Further, in more usefully decoupling the momentum and entropic contributions, the decomposition of Tij due to Lilley (Tech. Rep. AGARD CP-131 1974) is preferred. Connections are made between the present results and the quieting of high-speed jets with heating.

Two-dimensional Lagrangian acceleration statistics of inertial particles in a turbulent boundary layer with free-stream turbulence are determined by means of a particle tracking technique using a high-speed camera moving along the side of the wind tunnel at the mean flow speed. The boundary layer is formed above a flat plate placed horizontally in the tunnel, and water droplets are fed into the flow using two different methods: sprays placed downstream from an active grid, and tubes fed into the boundary layer from humidifiers. For the flow conditions studied, the sprays produce Stokes numbers varying from 0.47 to 1.2, and the humidifiers produce Stokes numbers varying from 0.035 to 0.25, where the low and high values refer to the outer boundary layer edge and the near-wall region, respectively. The Froude number is approximately 1.0 for the sprays and 0.25 for the humidifiers, with a small variation within the boundary layer. The free-stream turbulence is varied by operating the grid in the active mode as well as a passive mode (the latter behaves as a conventional grid). The boundary layer momentum-thickness Reynolds numbers are 840 and 725 for the active and passive grid respectively. At the outer edge of the boundary layer, where the shear is weak, the acceleration probability density functions are similar to those previously observed in isotropic turbulence for inertial particles. As the boundary layer plate is approached, the tails of the probability density functions narrow, become negatively skewed, and their peak occurs at negative accelerations (decelerations in the streamwise direction). The mean deceleration and its root mean square (r.m.s.) increase to large values close to the plate. These effects are more pronounced at higher Stokes number. In the vertical direction, there is a slight downward mean deceleration and its r.m.s., which is lower in magnitude than that of the streamwise component, peaks in the buffer region. Although there are free-stream turbulence effects, and the complex boundary layer structure plays an important role, a simple model suggests that the acceleration behaviour is dominated by shear, gravity and inertia. The results are contrasted with inertial particles in isotropic turbulence and with fluid particle acceleration statistics in a boundary layer. The background velocity field is documented by means of hot-wire anemometry and laser Doppler velocimetry measurements. These appear to be the first Lagrangian acceleration measurements of inertial particles in a shear flow.

We consider an infinite plate being withdrawn (at an angle α to the horizontal, with a constant velocity U) from an infinite pool of viscous liquid. Assuming that the effects of inertia and surface tension are weak, Derjaguin (C. R. Dokl. Acad. Sci. URSS, vol. 39, 1943, p. 13.) conjectured that the ‘load’ l, i.e. the thickness of the liquid film clinging to the plate, is l=(μU/ρgsinα)1/2, where ρ and μ are the liquid's density and viscosity, and g is the acceleration due to gravity.

In the present work, the above formula is derived from the Stokes equations in the limit of small slopes of the plate (without this assumption, the formula is invalid). It is shown that the problem has infinitely many steady solutions, all of which are stable – but only one of these corresponds to Derjaguin's formula. This particular steady solution can only be singled out by matching it to a self-similar solution describing the non-steady part of the film between the pool and the film's ‘tip’.

Even though the near-pool region where the steady state has been established expands with time, the upper, non-steady part of the film (with its thickness decreasing towards the tip) expands faster and, thus, occupies a larger portion of the plate. As a result, the mean thickness of the film is 1.5 times smaller than the load.

We investigate stability with respect to two-dimensional (independent of z) disturbances of plane-parallel shear flows with a velocity profile Vx=u(y) of a rather general form, monotonically growing upwards from zero at the bottom (y=0) to U0 as y → ∞ and having no inflection points, in an ideal incompressible fluid stably stratified in density in a layer of thickness ℓ, small as compared to the scale L of velocity variation. In terms of the ‘wavenumber k – bulk Richardson number J’ variables, the upper and lower (in J) boundaries of instability domains are found for each oscillation mode. It is shown that the total instability domain has a lower boundary which is convex downwards and is separated from the abscissa (k) axis by a strip of stability 0 < J < J0(−)(k) with minimum width J*=O(ℓ2/L2) at kL=O(1). In other words, the instability domain configuration is such that three-dimensional (oblique) disturbances are first to lose their stability when the density difference across the layer increases. Hence, in the class of flows under consideration, it is a three- not two-dimensional turbulence that develops as a result of primary instability.

The earlier constitutive model of Fang & Owens (Biorheology, vol. 43, 2006, p. 637) and Owens (J. Non-Newtonian Fluid Mech. vol. 140, 2006, p. 57) is extended in scope to include non-homogeneous flows of healthy human blood. Application is made to steady axisymmetric flow in rigid-walled tubes. The new model features stress-induced cell migration in narrow tubes and accurately predicts the Fåhraeus–Lindqvist effect whereby the apparent viscosity of healthy blood decreases as a function of tube diameter in sufficiently small vessels. That this is due to the development of a slippage layer of cell-depleted fluid near the vessel walls and a decrease in the tube haematocrit is demonstrated from the numerical results. Although clearly influential, the reduction in tube haematocrit observed in small-vessel blood flow (the so-called Fåhraeus effect) does not therefore entirely explain the Fåhraeus–Lindqvist effect.

Direct numerical simulation (DNS) of vortex shedding behind a tapered plate with the taper ratio 20 placed normal to the inflow has been performed. The Reynolds numbers based on the uniform inflow velocity and the width of the plate at the wide and narrow ends were 1000 and 250, respectively. For the first time ever cellular vortex shedding was observed behind a tapered plate in a numerical experiment (DNS). Multiple cells of constant shedding frequency were found along the span of the plate. This is in contrast to apparent lack of cellular vortex shedding found in the high-Reynolds-number experiments by Gaster & Ponsford (Aero. J., vol. 88, 1984, p. 206). However, the present DNS data is in good qualitative agreement with similar high-Reynolds-number experimental data produced by Castro & Watson (Exp. Fluids, vol. 37, 2004, p. 159). It was observed that a tapered plate creates longer formation length coupled with higher base pressure as compared to non-tapered (i.e. uniform) plates. The three-dimensional recirculation bubble was nearly conical in shape. A significant base pressure reduction towards the narrow end of the plate, which results in a corresponding increase in Strouhal number, was noticed. This observation is consistent with the experimental data of Castro & Rogers (Exp. Fluids, vol. 33, 2002, p. 66). Pressure-driven spanwise secondary motion was observed, both in the front stagnation zone and also in the wake, thereby reflecting the three-dimensionality induced by the tapering.